Changes in drinking status, serious illness and mortality

Objective: We studied the relationship of drinking and abstaining with all-cause mortality in a two-measurement-point prospective study to assess the importance of drinking change versus stability among the healthy and seriously ill. Method: The National Health and Nutrition Examination Survey I (NHANES I) and follow-up of men and women (age 22-49 and 50+) at first measurement (1971-74) who survived until second measurement (1982-84) were followed in death records for 10 years (1993). Cox proportional hazards models evaluated those with and without serious illness. Results: We found no elevated mortality risk for consistent never drinkers, but consistent heavier drinkers were at higher all-cause risk among men. Groups of abstainers significantly differed for mortality risk, and illness was implicated in these differences. New drinkers who formerly abstained did not improve their chances for longevity. Heavier drinking men who decreased consumption increased their chances for survival. Conclusions: Abstinence per se is not a risk factor for all-cause mortality, but heavier drinking is a risk factor among men. Adoption of drinking among former abstainers does not improve chances for survival. Reduction in consumption among heavier drinking men improves chances for survival.